OpenAI latest feature, ChatGPT, has the potential to detect students cheating
OpenAI is working on a new feature that could help catch students cheating using ChatGPT, a generative AI-based tool. The company has been working on a tool that can help users identify articles and essays written by AI for about a year. The tool has been a subject of internal debate for about two years and is ready to be released for about a year.
OpenAI is exploring text watermarking to track the origin of content, which is effective against localized tampering such as paraphrasing. However, it is less effective against globalized tampering, such as using translation systems, rewording with another generative model, or asking the model to insert a special character in between every word and then delete that character. This makes it trivial to circumvention by bad actors.
OpenAI latest feature is working on identifying text written by AI systems like ChatGPT.
In addition to text watermarking, OpenAI is also exploring using metadata to identify text written by AI or trace its origin. The company says that it is still in the early stages of exploring this technique but looks more promising than text watermarking. Metadata is cryptographically signed, meaning there are no false positives. While text watermarking has a low false positive rate, applying it to large volumes of text would lead to a large number of total false positives.
However, these tools remain unavailable for now, so there is no way to know if the essay that the students submitted has been written by AI. The latest tech and auto news from Techlusive can be found on their WhatsApp Channel, Facebook, X (Twitter), Instagram, and YouTube.
In summary, OpenAI is working on a new feature that could help catch students cheating using ChatGPT, which has increased its usage across various domains, including education. The company is also exploring the use of metadata to trace the origin of text, which is more promising than text watermarking.
Read These Articles Also :-
